Numeric ID as Categorical Variable

[pbaumgartner]

I have an variable that's an integer identifier. I want a chart that displays the average of some other variable for each unique option of this ID variable. When I try and do this in Power BI, it just displays the grand mean (or whatever other aggregate calculation) across each unique value, rather than each unique value's aggregation. How do I tell Power BI to treat this numeric ID Variable as a categorical?

[bidgeir · Accepted solution]

I might be misunderstanding something here but I have no problems doing this. Both the columns are whole numbers. All I have to do is put on column in Axis and the other in Value and Power BI does the rest. Are you doing something different?

[pbaumgartner]

That's exactly what I'm doing too, I even recreated with some different variables to ensure it wasn't just something weird about that specific combination.

If it makes any difference, I'm connected via ODBC to a Postgres database.

[Anonymous]

This can happen if table relationships are involved and not configured exactly right.  Are there any related tables?  Can you share sample data and/or your .PBIX?

...And just saw you got it fixed.  Good news.

[pbaumgartner]

Got it figured out, there was some artefact in the data that was causing the calculation execute incorrectly. Thanks for your help.